Search

Objectives : Oral conditions are of high prevalence and chronic character within the general population. Identifying the risk factors and determinants of oral disease is important, not only to reduce the burden of oral diseases, but also to improve (equal access to) oral health care systems, and to develop effective oral health promotion programs. Longitudinal population-based (birth-)cohort studies are very suitable to study risk factors on common oral diseases and have the potential to emphasize the importance of a healthy start for oral health. In this paper, we provide an overview of the comprehensive oral and craniofacial dataset that has been collected in the Generation R study: a population-based prospective birth cohort in the Netherlands that was designed to identify causes of health from fetal life until adulthood. Methods: Within the multidisciplinary context of the Generation R study, oral and craniofacial data has been collected from the age of 3 years onwards, and continued at the age of six, nine, and thirteen. Data collection is continuing in 17-year-old participants. Research outcomes: In total, the cohort population comprised 9749 children at birth, and 7405 eligible participants at the age of seventeen. Based on questionnaires, the dataset contains information on oral hygiene, dental visits, oral habits, oral health–related quality of life, orthodontic treatment, and obstructive sleep apnea. Based on direct measurements, the dataset contains information on dental caries, developmental defects of enamel, objective orthodontic treatment need, dental development, craniofacial characteristics, mandibular cortical thickness, and 3D facial measurements. Conclusions: Several research lines have been set up using the oral and craniofacial data linked with the extensive data collection that exists within the Generation R study. Clinical relevance: Being embedded in a multidisciplinary and longitudinal birth cohort study allows researchers to study several determinants of oral and craniofacial health, and to provide answers and insight into unknown etiologies and oral health problems in the general population. [ABSTRACT FROM AUTHOR]

Hashing learning is efficient for large-scale image retrieval by using the nearest neighbor search with binary codes instead of continuous representations. With the success of deep neural networks in related tasks such as data representation, recent hashing methods based on deep learning can further improve image retrieval quality and classification accuracy. However, most existing methods are primarily designed to maximize the performance of retrieval based on linear scan of hash codes which is still time-consuming on large-scale datasets. Fortunately, Hamming space retrieval is an alternative as it is less time-consuming by retrieving data points that are within a Hamming ball with a given Hamming radius, but few works focus on that. In this paper, we propose a concentrated hashing method with neighborhood embedding (CHNE) for efficient and effective image retrieval and classification. By integrating Cauchy cross-entropy and pair-wise weighted similarity loss, CHNE can enable similar data pairs with smaller Hamming distance and dissimilar data pairs with larger Hamming distance. In addition, existing hashing methods are usually designed for retrieval, thus the performance of classification using the binary codes is not guaranteed. To tackle this problem, we jointly minimize the regression quantization and neighborhood structure reconstruction errors in the loss function to improve the classification accuracy. The proposed end-to-end deep hashing method can be optimized by back-propagation in a standard manner. Experimental results on several datasets demonstrate that the proposed method can improve the performance of retrieval and classification. Due to its generality, the proposed method is expected to be useful for image retrieval and classification in broader areas. [ABSTRACT FROM AUTHOR]

The effluents of municipal wastewater treatment plant (WWTP) contain excessive nitrogen and phosphorus compared with the concentration in rivers or lakes. To reduce the pollutant load placed on aqueous environments, constructed wetland (CW) technology has been widely applied to advanced wastewater treatment. Packing substrates in CW could remove various pollutants. Steel slag, yellow earth, kaolin, volcanic rock, anthracite, and ceramsite could effectively remove phosphorus (P); volcanic rock, ceramsite, zeolite, yellow earth, manganese sand, and activated carbon have an affinity for ammonia nitrogen (NH4+-N). After 24 h reactions with the WWTP standard 1B synthetic wastewater, four packing substrates, i.e., volcanic rock and anthracite (1:1), volcanic rock and yellow earth (2:1), zeolite and yellow earth (2:1), and manganese sand and activated carbon (1:3), could remove over 56% and 30% of NH4+-N and phosphorus respectively. In addition, anthracite and volcanic rock (1:3), anthracite and activated carbon (1:40), anthracite and manganese sand (1:5), and anthracite and zeolite (1:4) effectively purified NH4+-N and phosphorus in secondary WWTP effluent, with removal efficiency exceeding 39% and 27%, respectively. A sequential experiment was performed to optimize packing substrates ratios in CW with volcanic rock and anthracite, ceramsite and yellow earth, and manganese sand and activated carbon. When the quantity of the substrate was doubled, most packing substrates adsorb more than 50% phosphorus and NH4+-N of the standard 1B WWTP synthetic wastewater. Considering the removal efficiency of packing substrates on phosphorus and NH4+-N, it is suggested that manganese sand and activated carbon (1:3), volcanic rock and anthracite (2:1), and yellow earth are appropriate substrates for CW in WWTP effluent advanced treatment. [ABSTRACT FROM AUTHOR]